The major effect of HIV-1 infection is elimination of the CD4+T cell population, the event that correlates with the profound immunosuppression caused by the virus. Antiretroviral therapy of the infection causes dramatic curtailment of virus replication, evidenced by virtual elimination of virus particles from plasma and recovery of CD4+T cell numbers. However, the virus persists in lymphoid tissues, and upon withdrawal of therapy (for whatever reason), virus replication begins again in most of these individuals, with renewed attack on the CD4+T cell population. The question arises whether a vaccine given during the period of drug therapy could induce protective immune responses that would prevent viral resurgence when the therapy is withdrawn. Problems with this concept are that immunological correlates of protection against HIV are not known, and a prophylactic vaccine, on which a therapeutic vaccine should be based, has not been developed yet. Even if an effective prophylactic vaccine were on hand, proof of concept for therapeutic immunization would still be required. In this application, we will use the SHIVKU/macaque model of HIV infection-disease to examine the feasibility of therapeutic immunization. SHIV has the env of HIV-1 on a background of SIV and is the closest genetically related agent to HIV-1 that will replicate productively in macaques. We adapted SHIV to macaques and derived a highly virulent strain, SHIVKU, which causes a very rapid type of HIV-like disease, with sub-total elimination of CD4+T cells and AIDS in 6 months. Using genetic engineering, we also recently developed a live virus vaccine that replicated transiently in macaques and induced immune responses that prevented replication of SHIVKU, given as challenge, for 2.5 years. Having established its efficacy prophylactically, we will now examine the potential of this vaccine for therapeutic efficacy. Macaques will be inoculated with SHIVKU then placed on temporary antiretroviral therapy, which will suppress replication of this virus. A portion of the animals will then be inoculated with drug-resistant vaccine virus. Following withdrawal of therapy, we will determine whether vaccine -induced immune responses inhibited resurgence of SHIVKU from reservoirs of latently infected cells in lymphoid tissues. In anticipation of success, we will then determine whether a more conservative therapeutic immunogen, inactivated virus particles, could also function therapeutically. These studies will establish proof-of- concept for the feasibility of performing this type of immunization to effectively prevent HIV resurgence in drug- treated infected individuals who come off therapy. This vaccine approach would provide a clear alternative to the dreary prospect of permanent post infection drug therapy.